Hi, Stefan,
 Had similar ?s as Piotr. Do you mean by 'algorithms' what may be termed 'spelling codes'? That is, are you looking for the quantity of different ways that people use combinations of letters in an alphabet to represent (write, spell) the particular speech sounds of spoken words that sound alike?
  Written English, besides transliterations from Greek, Arabic, etc. as you know, evolved from at least six main European languages that used variations of what people loosely call the Roman alphabet. About 30% of 'English' is from the French, which currently, is one of the worst alphabet languages that tries to represent spoken words with the archaic Roman alphabet. As a result, 'English' is extremely difficult for many who use it as a second language and for most of the people in the United States. Over forty million people in the U.S. are functionally illiterate. For the past 400 years no one ever knew what are the spelling codes of American English and shared that knowledge with the rest of us. I spent 36,000 hours attacking that problem since 1957. There are at least 1796 different codes for spelling the forty-two speech sounds of American.
Suspect the creators of PERL included a tiny fraction of them in its program.